What is Fabric Manager?

What is Fabric Manager?

FM configures the NVSwitch memory fabrics to form one memory fabric among all participating GPUs and monitors the NVLinks that support the fabric. At a high level, FM completes the following tasks:

• Configures routing (earlier than the fourth generation NVSwitch) among NVSwitch ports.

• Sets up GPU routing and port map if applicable.

• Coordinates with the GPU driver to initialize GPUs.

• Monitors the fabric for NVLink and NVSwitch errors.

• On systems that are not capable of Autonomous Link Initialization (ALI)-based NVLink training (first and second generation NVSwitch-based systems), FM complets the following tasks:

• Coordinates with the NVSwitch driver to initialize and train NVSwitch-to-NVSwitch NVLink interconnects.

• Coordinates with the GPU driver to initialize and train NVSwitch-to-GPU NVLink interconnects.

This user’s guide provides an overview of FM features and is intended for system administrators and NVSwitch-based server system users.